TETRACARBONYL HYDRIDE THEACID PROPERTIES O F IRON
Sept., 1949
3035
of optical density a t 222.5 mp versus p H shows clearly that a p H of ca. 5.0 is the optimum PH to use in acid solutions for this determination. For solutions of low ionic strength PH values of 5.0 * 1.5 are quite satisfactory, the optical density not varying appreciably in this region. Concentrations as low as lo-' molar in periodate should be detectable using 5.00-cm. silica cells. In the event that i t is desirable to determine periodate spectrophotometrically in alkaline solutions, p H values between ca. 10.5 and 12.5 are desirable. Such This molar ratio corresponds to PH values of 8.24 analysis leads to a sacrifice in sensitivity of the and 8.26 on curves 1 and 2, respectively, in Fig. 4. method a t 222.5 mp because of the lower molecular This gives a value of Kz = 5.6 X which, in extinction coefficient of the H3I06' ion as comview of the uncertainties involved in y1 and y2, is pared to that of the H4106- ion used in the lower in good agreement with the value of K2 = 4.35 X PH range. This disadvantage may be eliminated obtained from the spectrophotometric stud- by simply carrying out the analysis a t a somewhat ies. lower wave length, i. e., a t a wave length nearer to The value of Ks, also a new constant, agrees, or a t which the H3106" ion shows an absorption within experimental error, with the value of K3 maximum. obtained from K1 and Kz using the method of Hill.6 Summary Activity Coefficients.-The magnitudes of the It was shown that most probably any equilibria differences between Curves 1 and 2 in Fig. 3 are much greater than predicted differences due to involving the formation of the dimesoperiodate is of no importance in dilute aqueous such changes in ionic strength, using normal ion, 1209", y1 and y2 values. This suggests an abnormal de- solutions. Three simple equilibria involving stepcrease of tbe actual activity coefficients of the wise dissociation of paraperiodic acid, HJ06, acH4106- and H3106=ions, particularly the latter, count completely for the variation of the ultraviowith increasing ionic strength. For this reason let absorption spectra of dilute aqueous solutions values of K1 and K Zwere calculated from the data of periodate with PH. Kl4s5was substantiated and for low ionic strength where errors due to such ab- new estimates of K2 and K3 were made from specnormalities would be very small. Estimates of trophotometric studies. Kz was substantially the actual YI and YZ values a t the ionic strengths confirmed by potentiometric studies and K B was indicated may be made from the magnitudes of confirmed by the method of Hill.6 Evidence was shown for abnormal activity coefficients of H4106the differences between the two curves. Analytical Significance.-From these studies and H3IO6- ions with increasing ionic strength. i t is quite obvious that for the analytical deter- Recommendations were made for the spectrophomination of periodate (in the absence of inter- tometric determination of periodate in slightly fering substances) spectrophotometrically, p H is acid and alkaline solutions. a very important factor. Inspection of the plot AMES,IOWA RECEIVED APRIL4, 1949 Knowing the total quantity of periodate, the volume of solution, how the assumed values for 71 and ~2 vary with ionic strength, and how the ratio of Mz:1141 varies with ml. of base added, the condition where Eq. 7 is satisfied was evaluated by a series of successive approximations. This condition was found to occur where 4,Lwas ca. 0.25 and
'
[CONTRIBUTION FROM
THE
RESEARCH LABORATORY OF ORQUIMA s. A.,
si0 PAULO,BRAZIL]
The Acid Properties of Iron Tetracarbonyl Hydride BY P. KRUMHOLZ AND H. M. A. STETTINER
It has been shown previously' that the reaction between iron pentacarbonyl and alkalies does not produce, as assumed by HieberJ2the free iron tetracarbonyl hydride but its monobasic salts. The acid behavior of iron carbonyl hydride could be confirmed later by preparing salts with complex amine cation^.^ Some of them, as shown by Hieber,' behave in solution as strong electrolytes. Meanwhile Hieber' stated that structures as [Fe(C0)4H]- are not stable and only can be stabilized by the formation of salts (1) F. Feigl and P. Krumholz, Monalsh., 69, 314 (1932). (2) W.Hieber and F . Leutert, Z. anorg. Chcm., 204, 145 (1932). (3) F. Feigl and P. Krumholz, ibid., 416, 242 (19312 (4) W.Hieber and E. Fack, i b i d . , 286, 83 (1938).
with complex cations, that the alkaline solutions of the hydride contain aquo salts rather than normal salts not yet preparedIs and that the formation of metal derivatives of Fe(CO),HZ is b y far not as general as i t is supposed to be in an ordinary formation of salts. I n order to destroy any doubts regarding the behavior of iron carbonyl hydride as an ordinary acid we tried to determine its dissociation constant6 by potentiometric titration of its salts with (5) The monosodium salts prepared by F . Feigl and P. Krumholz. ref. 1, as alcoholates could not be freed completely from alcohol. (6) A. A. Blanchard, Chcm. Rev., 21, 3 (1937), assumed that the dissociation constant of cobalt tefracarbonyl hydride should be about lo-'.
3036
P. KKUMHVLZ AND H. M. A. STETTINER
V d , 71
it corresponds to a thermodynamic dissociation (at 17.5’). Computing constant’ of -3.4 X the dissociation constant from p H values a t different concentrations of [Fe(C0)4H]-, and assuming that the concentration of undissociated Fe(C0)dHz remains constant, we obtained as mean value K I = 4.15 X (Table I).
8
6
TABLEI
z
fiH
CF~(CO)&E
Q,
1 x 10-1 7 . 5 x 10-2 5 . 3 x 10-2 4 . 3 x 10-2 2 . 3 X 10-2 8 . 4 X 10-3
4
2
10 20 30 HC1, cc. Fig. 1.-Titration of 0.055 m. [Fe(CO)dH]zBa with 0.57 N HC1; initial volume 135 cc., T = 17.5’.
K1
6.00 5.88 5.71 5.63 5.36 4.95
x
105
4.17 4.13 4.31 4.20 4.14 3.92
The titration curve shows furthermore that the reaction between barium hydroxide and excess Fe(C0)5 proceeds to an extent of at least 99.8% according to (1). Solutions of the mono potassium salt of iron carbonyl hydride prepared by allowing Fe(C0)5 to react with potassium and barium hydroxide in quantities corresponding to the equation Fe(C0)6 f KOH + Ba(0H)p = Fe(C0)4KH BaCOs
+ + H20
(2)
behave in a similar way. As iron carbonyl hydride also forms bibasic salts with some heavy metals,***and as Blanchards recently assumed the formation of a bibasic potassium salt Fe(C0)4K2, we submitted solutions prepared according to Fe(C0)5
I
I
1 2 Concn. Fe(CO)4H~X lo3. Fig. 2.-Titration of 0.055 m. [Fe(CO)rH]zBawith 0.57 N HC1.
HC1. Figs. 1 and 2 show the change of pH during the titration of a solution of the barium salt [Fe(C0)4H]2Baobtained by allowing barium hydroxide to react upon an excess of iron pentacarbonyl’ according to 2Fe(CO)b 4-3Ba(OH)z = [Fe(CO)dH]2Ba 2BaC03
+ + 2H2O
(1)
Those diagrams represent typical titration curves of salts of a weak acid of low solubility, since, immediately after starting the titration, the initially steep slope of the titration curve turns suddenly flat. At this point the initially clear solution becomes turbid due to the separation of free iron carbonyl hydride (the point where the first visible turbidity appears is indicated by an arrow in Fig. 2). From this part of the titration curve the molar solubility of Fe(CO)*H2 (at 17‘6’) may be calculated as about 1.8 X IO+. The $44 at the endpoint at the titration is 8.60;
+ 2KOH + Ba(OH)2
=
+
Fe(C0)4K2 BaCO3 2H20
+
(3)
to potentiometric titrations. As shown in Fig. 3 the p H values in the first moiety of the diagram are lower than the corresponding values of potassium hydroxide of equal concentration (represented by the dotted line in Fig. 3). This proves the formation of the dibasic salt, hydrolyzed to an extent of about 457’ in 0.18 molar solution. We computed the second dissociation constant of iron carbonyl hydride from the hydrolytic equilibrium of the bibasic salt according t o lo where CK,X and CKHXrepresent the analytical concentrations of the bibasic and monobasic salt, K H ~ O 5.6 X (at 17.5’) and y the activity coefficients of the ions. l1 We obtained (7) Assuming the activity coefficient OF [Fe(CO)
